Configurable parameters for pure pursuit

Visualize given trajectory

Run the tracker simulation

  1. Instantiate the tracker class
  2. Initialize some starting pose
  3. Simulate robot motion 1 step at a time - get $v$, $\omega$ from tracker, predict new pose using $v$, $\omega$, current pose in simulate_unicycle()
  4. Stop simulation if tracker declares that end-of-path is reached
  5. Record all parameters

Visualize curvature

Animate

Make a video to plot the current pose of the robot and reference pose it is trying to track. You can use funcAnimation in matplotlib

Effect of noise in simulations

What happens if you add a bit of Gaussian noise to the simulate_unicycle() output? Is the tracker still robust?

The noise signifies that $v$, $\omega$ commands did not get realized exactly